Retargeting Funnel

SKILL.md

---
name: Retargeting Funnel
description: Designs multi-stage retargeting ad funnels that re-engage ecommerce visitors based on browsing depth — from homepage bouncers to cart abandoners — with platform-specific strategies for Meta, Google, and TikTok.
---

# Retargeting Funnel

Design and deploy multi-stage retargeting ad funnels that bring ecommerce visitors back at every level of buying intent, matching creative format, bid strategy, and frequency cap to each audience segment's proximity to purchase.

## Quick Reference

| Decision | Strong | Acceptable | Weak |
|---|---|---|---|
| Audience window for cart abandoners | 1-3 days with urgency creative | 3-7 days with reminder creative | 14+ days with generic ads |
| Product page viewer retargeting | Dynamic product ads showing viewed items | Category-level carousel ads | Static brand awareness ads |
| Homepage bouncer strategy | Broad value-prop video ads (15s) | Lifestyle image carousel | Same ads as cart abandoners |
| Frequency capping | 3-5 impressions per day per segment | 6-8 impressions per day | Uncapped or 15+ per day |
| Cross-platform sequencing | Platform-specific creative per stage | Same creative resized per platform | Single platform only |
| Exclusion logic | Exclude purchasers + higher-funnel segments from lower-funnel campaigns | Exclude purchasers only | No exclusions between segments |
| Budget allocation by funnel stage | 50-60% bottom funnel, 25-30% mid, 15-20% top | Even split across stages | Majority spend on top funnel |
| Lookback window overlap handling | Strict non-overlapping windows with priority rules | Partial overlap with bid adjustments | Fully overlapping audiences |

## Solves

1. **Wasted retargeting spend on low-intent visitors** — Segments audiences by browsing depth so budget flows to the visitors most likely to convert rather than blanketing everyone with the same ad.
2. **Cart abandonment revenue leakage** — Deploys time-sensitive, product-specific creative within hours of abandonment to recover 8-15% of lost carts.
3. **Ad fatigue from undifferentiated retargeting** — Rotates creative themes and formats across funnel stages so each visitor sees progressively relevant messaging instead of the same banner repeatedly.
4. **Cross-platform attribution confusion** — Establishes clear platform roles (Meta for discovery retargeting, Google for intent capture, TikTok for social proof) with unified audience definitions.
5. **Over-retargeting recent purchasers** — Implements exclusion audiences and suppression windows to prevent showing ads to customers who already converted.
6. **Inability to measure incremental lift from retargeting** — Builds holdout groups and incrementality tests into the funnel structure so you can prove retargeting is driving real revenue, not just claiming organic conversions.
7. **Missing the mid-funnel entirely** — Captures the product-page-viewer and collection-browser segments that most advertisers ignore, filling the gap between awareness and cart-level intent.

## Workflow

### Step 1: Audit Existing Pixel and Event Setup

Before building any audiences, verify that your tracking infrastructure captures the events needed for segmentation.

**Required events by platform:**
- Meta Pixel: PageView, ViewContent, AddToCart, InitiateCheckout, Purchase
- Google Ads: page_view, view_item, add_to_cart, begin_checkout, purchase
- TikTok Pixel: PageView, ViewContent, AddToCart, InitiateCheckout, CompletePayment

**Validation checklist:**
- Fire each event manually and confirm it appears in the platform's event manager
- Verify event parameters include product ID, value, currency, and content category
- Check that the pixel fires on all pages, including dynamic product pages loaded via JavaScript
- Confirm server-side event forwarding (CAPI for Meta, enhanced conversions for Google) is active to offset browser-side tracking losses

**Output:** A pixel health report listing each event, its fire rate over the past 7 days, and any parameter gaps.

### Step 2: Define Audience Segments by Funnel Stage

Create mutually exclusive audience segments based on the deepest action each visitor took. Exclusion logic is critical — a cart abandoner should not also appear in your homepage bouncer audience.

| Segment | Definition | Lookback Window | Typical Size (% of traffic) |
|---|---|---|---|
| S1: Homepage Bouncers | Visited site, viewed no product pages | 30 days | 25-40% |
| S2: Category Browsers | Viewed collection/category pages but no product detail pages | 21 days | 15-25% |
| S3: Product Viewers | Viewed 1+ product detail pages, did not add to cart | 14 days | 15-25% |
| S4: Cart Abandoners | Added 1+ items to cart, did not begin checkout | 7 days | 5-10% |
| S5: Checkout Abandoners | Began checkout, did not purchase | 3 days | 2-5% |
| S6: Recent Purchasers (suppress/upsell) | Completed purchase | 14-30 days | 2-5% |

**Exclusion hierarchy:** Each segment excludes all segments below it. S1 excludes S2-S6. S4 excludes S5-S6. This prevents audience overlap and ensures budget is not wasted showing low-intent creative to high-intent users.

### Step 3: Design Creative by Funnel Stage

Match ad format, messaging angle, and call-to-action intensity to the visitor's demonstrated intent level.

**S1 — Homepage Bouncers:**
- Format: 15-second video ad or lifestyle carousel
- Message: Brand story, unique value proposition, social proof (press mentions, customer count)
- CTA: "Discover Our Collection" or "See What's New"
- Goal: Drive them back to browse

**S2 — Category Browsers:**
- Format: Collection-level carousel or dynamic category ads
- Message: Highlight the category they browsed, show bestsellers within it
- CTA: "Shop [Category Name]" or "Trending in [Category]"
- Goal: Move them to product pages

**S3 — Product Viewers:**
- Format: Dynamic product ads (DPA) showing exact viewed items + similar products
- Message: Product benefits, ratings, limited stock signals
- CTA: "Still Interested?" or "Back in Stock"
- Goal: Get them to add to cart

**S4 — Cart Abandoners:**
- Format: Dynamic ads showing carted items with price
- Message: "You left something behind" with urgency (limited stock, sale ending)
- CTA: "Complete Your Order" with direct cart link
- Goal: Return to cart and purchase

**S5 — Checkout Abandoners:**
- Format: Single-product ad showing the primary carted item
- Message: "Your order is waiting" — address objections (free shipping, easy returns, secure payment)
- CTA: "Finish Checkout" with incentive if margin allows (5-10% discount, free shipping code)
- Goal: Complete purchase

### Step 4: Configure Platform-Specific Campaigns

Set up campaigns on each platform with the right objective, bid strategy, and placement.

**Meta Ads:**
- Use Advantage+ catalog sales for S3-S5 (dynamic product ads)
- Use Traffic or Engagement objective for S1-S2
- Enable Advantage+ placements but review delivery by placement weekly
- Set campaign budget optimization (CBO) within each funnel stage, not across stages

**Google Ads:**
- Performance Max with audience signals for S3-S5
- Standard Display with custom intent audiences for S1-S2
- YouTube retargeting for S1 with 15-second bumper ads
- Set target ROAS bidding for bottom-funnel, maximize clicks for top-funnel

**TikTok Ads:**
- Spark Ads using organic content for S1-S2 (highest engagement rate)
- Collection Ads for S3 showing viewed products
- Conversion objective with value optimization for S4-S5
- Use TikTok's auto-creative optimization for format testing

### Step 5: Set Frequency Caps and Budget Allocation

**Frequency caps by segment:**

| Segment | Daily Cap | Weekly Cap | Rationale |
|---|---|---|---|
| S1 | 2-3 | 10-12 | Low intent — avoid annoyance |
| S2 | 3-4 | 14-16 | Building familiarity |
| S3 | 4-5 | 18-20 | High intent — stay visible |
| S4 | 5-6 | 20-25 | Urgency window is short |
| S5 | 5-7 | 25-30 | Highest conversion probability |

**Budget allocation framework:**

Allocate budget proportional to expected return, not audience size. Bottom-funnel segments are smaller but convert at 5-15x the rate of top-funnel.

- S5 Checkout Abandoners: 20-25% of retargeting budget
- S4 Cart Abandoners: 25-30%
- S3 Product Viewers: 20-25%
- S2 Category Browsers: 10-15%
- S1 Homepage Bouncers: 5-10%

### Step 6: Build Measurement and Optimization Framework

**Key metrics by funnel stage:**

| Segment | Primary Metric | Target Benchmark | Secondary Metric |
|---|---|---|---|
| S1 | Click-through rate | 0.8-1.5% | Cost per site visit |
| S2 | Product page view rate | 15-25% of clickers | Cost per product view |
| S3 | Add-to-cart rate | 8-15% of clickers | Cost per add-to-cart |
| S4 | Purchase conversion rate | 10-20% | Cost per acquisition |
| S5 | Purchase conversion rate | 15-30% | Return on ad spend |

**Incrementality testing:**
- Hold out 10-15% of each segment from retargeting for 2-4 weeks
- Compare conversion rate of exposed group vs. holdout
- Calculate incremental ROAS = (revenue from exposed - revenue from holdout) / ad spend
- If incremental lift is below 5% for a segment, reallocate its budget to higher-performing stages

### Step 7: Iterate with Cohort Analysis

Run weekly cohort reviews to identify shifts in funnel behavior:
- Track segment migration rates (what % of S1 visitors become S3 within 7 days)
- Monitor time-to-conversion by segment to adjust lookback windows
- A/B test creative rotations within each stage on a 2-week cycle
- Refresh dynamic product feed data daily to prevent stale product images or prices
- Sunset audiences that exceed their lookback window by adding rolling exclusions

## Worked Examples

### Example 1: DTC Skincare Brand ($150K/month ad spend)

**Context:** A direct-to-consumer skincare brand selling $40-$120 products wants to improve retargeting efficiency. Current retargeting runs a single campaign targeting all site visitors from the past 30 days with the same carousel ad. Current ROAS on retargeting: 2.8x.

**Step 1 — Pixel Audit:**
Meta Pixel and Google tag are both active. TikTok pixel is installed but AddToCart event is not firing on the AJAX cart — requires developer fix. Meta CAPI is configured. Google enhanced conversions are not set up (flagged for implementation).

**Step 2 — Segments Built:**

| Segment | 30-Day Size | Lookback | Exclusions |
|---|---|---|---|
| S1: Homepage Bouncers | 85,000 | 30 days | Excludes S2-S6 |
| S2: Category Browsers | 42,000 | 21 days | Excludes S3-S6 |
| S3: Product Viewers | 38,000 | 14 days | Excludes S4-S6 |
| S4: Cart Abandoners | 12,000 | 7 days | Excludes S5-S6 |
| S5: Checkout Abandoners | 4,500 | 3 days | Excludes S6 |
| S6: Purchasers (suppress) | 6,200 | 30 days | — |

**Step 3 — Creative Plan:**
- S1: 15-second founder story video ("Why I started this brand") + press mention carousel
- S2: "Best of [Category]" carousel — top 5 products in browsed category with star ratings
- S3: DPA showing viewed products + "Customers Also Loved" companion products
- S4: "Still in your cart" DPA with free shipping reminder (orders over $75)
- S5: Single-product ad with 10% discount code, urgency copy ("Your cart expires soon")

**Step 4 — Platform Setup:**
- Meta: 3 campaign groups (Top: S1-S2 / Mid: S3 / Bottom: S4-S5), CBO within each
- Google: Performance Max for S3-S5, Standard Display for S1-S2
- TikTok: Paused until AddToCart pixel fix ships; planned for S1-S2 Spark Ads

**Step 5 — Budget Allocation (of $45K retargeting budget, 30% of total):**
- S5: $11,250 (25%)
- S4: $13,500 (30%)
- S3: $9,000 (20%)
- S2: $6,750 (15%)
- S1: $4,500 (10%)

**Step 6 — Results after 8 weeks:**

| Segment | ROAS | CPA | Incremental Lift |
|---|---|---|---|
| S5 | 9.2x | $8.40 | 22% over holdout |
| S4 | 6.8x | $12.10 | 18% over holdout |
| S3 | 4.1x | $22.50 | 14% over holdout |
| S2 | 2.3x | $38.00 | 7% over holdout |
| S1 | 1.4x | $52.00 | 3% over holdout |

**Optimization decisions:**
- Reduced S1 budget to 5%, reallocated to S4 (highest incremental return per dollar)
- S2 creative refreshed — swapped static images for UGC video, CTR improved 40%
- Overall retargeting ROAS improved from 2.8x to 5.1x

---

### Example 2: Multi-Brand Fashion Retailer ($500K/month ad spend)

**Context:** An online fashion retailer selling 200+ brands at $50-$400 price points. High traffic volume (2M monthly visitors) but retargeting campaigns are poorly segmented — one campaign for "all visitors" and one for "cart abandoners." Retargeting represents 35% of total spend ($175K/month). Current blended retargeting ROAS: 3.2x.

**Step 1 — Pixel Audit:**
All platforms (Meta, Google, TikTok) have pixels with complete event coverage. However, product feed has 1,200 items with missing GTIN numbers and 340 items with outdated prices. Feed issues are causing DPA disapprovals. Immediate fix: sync product feed from Shopify every 4 hours instead of daily.

**Step 2 — Segments Built (with sub-segments for this volume):**

| Segment | 30-Day Size | Lookback | Special Rules |
|---|---|---|---|
| S1: Homepage Bouncers | 480,000 | 21 days | Split by traffic source (paid vs. organic) |
| S2: Category Browsers | 320,000 | 14 days | Split by gender category browsed |
| S3a: Single Product Viewers | 280,000 | 14 days | Viewed 1 product |
| S3b: Multi-Product Viewers | 145,000 | 14 days | Viewed 3+ products (higher intent) |
| S4: Cart Abandoners | 68,000 | 7 days | Split by cart value (<$100 vs. $100+) |
| S5: Checkout Abandoners | 22,000 | 3 days | All treated as highest priority |
| S6: Purchasers | 45,000 | 60 days | 0-14 days: suppress; 14-60 days: cross-sell |

**Step 3 — Creative Plan:**
- S1 (paid traffic): "New Arrivals This Week" video lookbook — recapture the interest that brought them via paid ads
- S1 (organic): Brand manifesto + bestseller carousel — build brand affinity
- S2: Gender-specific trend roundups ("Men's Spring Essentials" / "Women's Date Night Edit")
- S3a: DPA of viewed product + 3 similar items at comparable price points
- S3b: "Your Shortlist" DPA showing all viewed items in a single carousel — acknowledge their research
- S4 (<$100 cart): "Don't miss out" with social proof ("2,300 people bought this today")
- S4 ($100+ cart): Free expedited shipping offer to justify the larger purchase
- S5: "Complete your order" with live chat CTA for objection handling + 48-hour price guarantee

**Step 4 — Platform Distribution:**
- Meta (60% of retargeting spend): All segments active. Advantage+ catalog for S3-S5. Video-first for S1-S2.
- Google (30%): Performance Max for S3-S5 with product-level ROAS targets. YouTube pre-roll for S1. Discovery ads for S2.
- TikTok (10%): S1-S2 only using Spark Ads from brand creator partnerships. Testing S3 with Collection Ads.

**Step 5 — Budget Allocation ($175K retargeting):**
- S5: $35,000 (20%)
- S4: $52,500 (30%)
- S3b: $22,750 (13%)
- S3a: $17,500 (10%)
- S2: $26,250 (15%)
- S1: $21,000 (12%)

**Step 6 — Results after 12 weeks:**

| Segment | ROAS | Conv. Rate | Cost Per Purchase |
|---|---|---|---|
| S5 | 11.4x | 24% | $11.20 |
| S4 ($100+) | 8.9x | 18% | $14.80 |
| S4 (<$100) | 5.6x | 15% | $9.60 |
| S3b | 5.2x | 9% | $18.40 |
| S3a | 3.4x | 5% | $28.90 |
| S2 | 2.1x | 2.2% | $42.00 |
| S1 (paid) | 1.6x | 0.9% | $58.00 |
| S1 (organic) | 1.1x | 0.5% | $71.00 |

**Optimization decisions:**
- Killed S1 organic retargeting (incremental lift only 1.2%) — reallocated to S4
- Doubled down on S3b multi-product viewers with higher frequency cap — this segment showed strongest incrementality (19%)
- Introduced cross-sell campaign for S6 purchasers at day 14 post-purchase, generating an additional $28K/month
- Blended retargeting ROAS improved from 3.2x to 5.8x over 12 weeks

## Common Mistakes

1. **Treating all site visitors as one retargeting audience.** A homepage bouncer has fundamentally different intent than a checkout abandoner. Showing them the same ad wastes budget on low-intent visitors and under-serves high-intent ones. Always segment by browsing depth.

2. **Not excluding higher-intent segments from lower-intent campaigns.** Without exclusion logic, a cart abandoner receives both the cart-specific ad and the generic "come visit us" ad. This dilutes messaging and inflates frequency. Build strict exclusion hierarchies.

3. **Setting identical lookback windows for all segments.** A 30-day window for checkout abandoners is wasteful — most who will convert do so within 72 hours. Shorter windows for higher-intent segments improve ROAS and reduce ad fatigue.

4. **Ignoring frequency caps entirely.** Retargeting without frequency limits causes ad fatigue, brand damage, and wasted impressions. Users who see the same ad 20+ times per week develop negative brand associations. Set caps and monitor them weekly.

5. **Allocating budget proportional to audience size instead of conversion probability.** Homepage bouncers are your largest audience but lowest-converting. Cart and checkout abandoners are tiny but convert at 10-20x the rate. Weight budget toward the bottom of the funnel.

6. **Using static creative for dynamic-eligible segments.** If a visitor looked at specific products, show them those exact products with dynamic product ads. Generic brand images for product viewers and cart abandoners leave conversion rate on the table.

7. **Running retargeting without incrementality measurement.** Without holdout tests, you cannot distinguish between conversions retargeting caused and conversions that would have happened organically. Many retargeting campaigns show high ROAS but near-zero incrementality on top-funnel segments.

8. **Neglecting product feed quality.** Dynamic product ads are only as good as your product feed. Outdated prices, missing images, out-of-stock items, and broken URLs in the feed erode user trust and trigger ad disapprovals. Sync feeds at minimum every 6 hours.

9. **Forgetting to suppress recent purchasers.** Showing "buy this product" ads to someone who bought it yesterday is a poor customer experience and wasted spend. Suppress purchasers for at least 14 days (longer for high-AOV or infrequent-purchase categories).

10. **Running the same creative for more than 3-4 weeks without rotation.** Even strong creative fatigues. Monitor frequency vs. CTR curves and rotate creative when CTR drops 20% below its initial benchmark.

## Resources

- [Meta Ads Retargeting Best Practices](https://www.facebook.com/business/help/retargeting) — Official Meta documentation on custom audiences, dynamic ads, and catalog campaigns
- [Google Ads Remarketing Guide](https://support.google.com/google-ads/answer/2453998) — Setup and optimization for Google Display, YouTube, and Performance Max remarketing
- [TikTok Ads Manager Retargeting](https://ads.tiktok.com/help/) — TikTok pixel events, custom audiences, and Spark Ads setup
- [Shopify Retargeting Audiences](https://help.shopify.com/en/manual/promoting-marketing/create-marketing/facebook-audiences) — Shopify-specific audience sync for Meta and Google
- [Dynamic Product Ads Feed Specification](https://www.facebook.com/business/help/120325381656392) — Meta catalog feed requirements for DPA campaigns
- [Google Merchant Center Feed Guide](https://support.google.com/merchants/answer/7052112) — Product feed specifications for Google Shopping and Performance Max
- [Incrementality Testing Framework](https://www.facebook.com/business/help/1738164643098669) — Meta conversion lift studies and holdout group configuration